FAQ Search Today's Posts Mark Forums Read
» Video Reviews

» Linux Archive

Linux-archive is a website aiming to archive linux email lists and to make them easily accessible for linux users/developers.


» Sponsor

» Partners

» Sponsor

Go Back   Linux Archive > Debian > Debian GCC

 
 
LinkBack Thread Tools
 
Old 05-07-2012, 05:06 PM
Julien Cristau
 
Default Accepted gcc-defaults 1.118 (source all amd64)

On Mon, May 7, 2012 at 16:19:00 +0000, Matthias Klose wrote:

> gcc-defaults (1.118) unstable; urgency=low
> .
> * Default to GCC 4.7 for gcc, g++, gfortran on amd64, i386,
> kfreebsd-amd64, kfreebsd-i386 and hurd-i386.

Please revert this change. There are still too many open bugs on
http://bugs.debian.org/cgi-bin/pkgreport.cgi?users=debian-gcc@lists.debian.org;tag=ftbfs-gcc-4.7,
and it's too late in the release cycle, to change this, IMO.

Cheers,
Julien
 
Old 05-07-2012, 06:19 PM
Cyril Brulebois
 
Default Accepted gcc-defaults 1.118 (source all amd64)

Julien Cristau <jcristau@debian.org> (07/05/2012):
> On Mon, May 7, 2012 at 16:19:00 +0000, Matthias Klose wrote:
>
> > gcc-defaults (1.118) unstable; urgency=low
> > .
> > * Default to GCC 4.7 for gcc, g++, gfortran on amd64, i386,
> > kfreebsd-amd64, kfreebsd-i386 and hurd-i386.
>
> Please revert this change. There are still too many open bugs on
> http://bugs.debian.org/cgi-bin/pkgreport.cgi?users=debian-gcc@lists.debian.org;tag=ftbfs-gcc-4.7,
> […]

Seconded. Furthermore, doing so without an ACK from the release team
isn't very nice. Too many transitions going on/to do already, we don't
need that kind of fun. Why do we need to take that extra blast?

Mraw,
KiBi.
 
Old 05-07-2012, 06:58 PM
Matthias Klose
 
Default Accepted gcc-defaults 1.118 (source all amd64)

On 07.05.2012 20:19, Cyril Brulebois wrote:
> Julien Cristau <jcristau@debian.org> (07/05/2012):
>> On Mon, May 7, 2012 at 16:19:00 +0000, Matthias Klose wrote:
>>
>>> gcc-defaults (1.118) unstable; urgency=low . * Default to GCC 4.7 for
>>> gcc, g++, gfortran on amd64, i386, kfreebsd-amd64, kfreebsd-i386 and
>>> hurd-i386.
>>
>> Please revert this change. There are still too many open bugs on
>> http://bugs.debian.org/cgi-bin/pkgreport.cgi?users=debian-gcc@lists.debian.org;tag=ftbfs-gcc-4.7,
>>
>>
[…]
>
> Seconded. Furthermore, doing so without an ACK from the release team isn't
> very nice. Too many transitions going on/to do already, we don't need that
> kind of fun. Why do we need to take that extra blast?

see my email to -release. It's ahead of the freeze, and the fix rate for the
build failures shows that all these issues can be fixed even before the
freeze. It is not a transition which requires binNMU's itself (except for the
libobjc stuff), and shouldn't affect other ongoing transitions. If there are
fixes needed for an ongoing transition, please bring these to my attention.

Matthias


--
To UNSUBSCRIBE, email to debian-gcc-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 4FA81B6A.8000806@debian.org">http://lists.debian.org/4FA81B6A.8000806@debian.org
 
Old 05-08-2012, 12:44 PM
Cyril Brulebois
 
Default Accepted gcc-defaults 1.118 (source all amd64)

Matthias Klose <doko@debian.org> (07/05/2012):
> see my email to -release. It's ahead of the freeze, and the fix rate for the
> build failures shows that all these issues can be fixed even before the
> freeze. It is not a transition which requires binNMU's itself (except for the
> libobjc stuff), and shouldn't affect other ongoing transitions.

Wrong. At least strigi, gnome-commander, and hugin are getting hit.

What's worse is that at least for the last two, there were no gcc 4.7
bug reports for them! And don't tell me thins changed for them, since
hugin is getting his *fourth* binNMU in a row.

You're making our job difficult. It's really ***NOT*** appreciated.

Mraw,
KiBi.
 
Old 05-08-2012, 04:55 PM
Matthias Klose
 
Default Accepted gcc-defaults 1.118 (source all amd64)

On 08.05.2012 14:44, Cyril Brulebois wrote:
> Matthias Klose <doko@debian.org> (07/05/2012):
>> see my email to -release. It's ahead of the freeze, and the fix rate for
>> the build failures shows that all these issues can be fixed even before
>> the freeze. It is not a transition which requires binNMU's itself (except
>> for the libobjc stuff), and shouldn't affect other ongoing transitions.
>
> Wrong. At least strigi, gnome-commander, and hugin are getting hit.

stop blaming the compiler for unrelated build failures. You did so already
yesterday:

<mbiebl> awesome, libchamplain ftbfs with an ICE in gcc-4.7
<mbiebl> yeah for switching the default compiler
<KiBi> doko: It looks like I'm going to take up your offer. Please fix that.
<mbiebl> doko: you have your first problem of gcc-4.7 breaking an ongoing
transition
<adsb> mbiebl: that might be the gmp10 bug
<doko> KiBi, can't reproduce the libchamplain ICE locally. could you do a
local build as well?

and I didn't get a reply on that.

strigi was fixed, the gnome-commander and hugin builds were both tried on
brahms (and did succeed on other architectures).

> What's worse is that at least for the last two, there were no gcc 4.7 bug
> reports for them! And don't tell me thins changed for them, since hugin is
> getting his *fourth* binNMU in a row.

no, because it's a buildd issue, or an gmp issue, afaics. Again, please don't
blame the compiler if you can't explain a build failure.

so what is the status of brahms?

> You're making our job difficult. It's really ***NOT*** appreciated.

well, not deciding on the release architectures for over a year now makes my
job difficult as well. still trying to get it done.

Matthias


--
To UNSUBSCRIBE, email to debian-gcc-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 4FA95015.4080003@debian.org">http://lists.debian.org/4FA95015.4080003@debian.org
 
Old 05-08-2012, 05:06 PM
Julien Cristau
 
Default Accepted gcc-defaults 1.118 (source all amd64)

On Tue, May 8, 2012 at 18:55:49 +0200, Matthias Klose wrote:

> strigi was fixed, the gnome-commander and hugin builds were both tried on
> brahms (and did succeed on other architectures).
>
Other architectures didn't switch to 4.7.

> > What's worse is that at least for the last two, there were no gcc 4.7 bug
> > reports for them! And don't tell me thins changed for them, since hugin is
> > getting his *fourth* binNMU in a row.
>
> no, because it's a buildd issue, or an gmp issue, afaics. Again, please don't
> blame the compiler if you can't explain a build failure.
>
No it ain't.

> so what is the status of brahms?
>
It's happy since it reverted to gmp 5.0.4. Less so since it upgraded
gcc to 4.7.

> > You're making our job difficult. It's really ***NOT*** appreciated.
>
> well, not deciding on the release architectures for over a year now makes my
> job difficult as well. still trying to get it done.
>
If you have specific concerns with a specific arch feel free to raise
them on -release and the corresponding port's mailing list, and we'll
take that into account. Not in this thread.

Cheers,
Julien
 
Old 05-08-2012, 05:08 PM
Julien Cristau
 
Default Accepted gcc-defaults 1.118 (source all amd64)

On Tue, May 8, 2012 at 18:55:49 +0200, Matthias Klose wrote:

> <mbiebl> awesome, libchamplain ftbfs with an ICE in gcc-4.7
> <mbiebl> yeah for switching the default compiler
> <KiBi> doko: It looks like I'm going to take up your offer. Please fix that.
> <mbiebl> doko: you have your first problem of gcc-4.7 breaking an ongoing
> transition
> <adsb> mbiebl: that might be the gmp10 bug
> <doko> KiBi, can't reproduce the libchamplain ICE locally. could you do a
> local build as well?
>
> and I didn't get a reply on that.
>
That was the gmp bug.

Cheers,
Julien
 
Old 05-08-2012, 05:14 PM
Cyril Brulebois
 
Default Accepted gcc-defaults 1.118 (source all amd64)

Matthias Klose <doko@debian.org> (08/05/2012):
> <adsb> mbiebl: that might be the gmp10 bug
> <doko> KiBi, can't reproduce the libchamplain ICE locally. could you do a
> local build as well?
>
> and I didn't get a reply on that.

Because, if you didn't get the memo, we're trying to get packages built,
which we finally managed to get. Of course reading some more lines would
have been too hard.

> strigi was fixed, the gnome-commander and hugin builds were both tried on
> brahms (and did succeed on other architectures).

Strigi needed many fixes, with a maintainer who kindly did everything
that was needed, and there were changes needed *BECAUSE OF YOU*. Read
the fucking changelog.

Gnome-commander needs a gcc 4.7 patch. Read the fucking bug report.

Hugin needs a gcc 4.7 patch. Read the fucking bug report.

> no, because it's a buildd issue, or an gmp issue, afaics. Again,
> please don't blame the compiler if you can't explain a build failure.

What you can see is bullshit. Read the fucking bugs, and patches.

> so what is the status of brahms?

Irrelevant.

> well, not deciding on the release architectures for over a year now
> makes my job difficult as well. still trying to get it done.

Changing the default compiler unilaterally, and with no prior notice,
and with no care about the ongoing needs as far as rebuilds and
transitions are concerned, is totally insane and makes our job HARDER.

Please stop being a pain in the ass. Thanks already.

Mraw,
KiBi.
 
Old 05-08-2012, 05:22 PM
"Adam D. Barratt"
 
Default Accepted gcc-defaults 1.118 (source all amd64)

On 08.05.2012 17:55, Matthias Klose wrote:
stop blaming the compiler for unrelated build failures. You did so
already

yesterday:

<mbiebl> awesome, libchamplain ftbfs with an ICE in gcc-4.7
<mbiebl> yeah for switching the default compiler
<KiBi> doko: It looks like I'm going to take up your offer. Please
fix that.
<mbiebl> doko: you have your first problem of gcc-4.7 breaking an
ongoing

transition
<adsb> mbiebl: that might be the gmp10 bug
<doko> KiBi, can't reproduce the libchamplain ICE locally. could you
do a

local build as well?

and I didn't get a reply on that.


Was that last line also in #-release? I (predictably) recognise the
rest, but the message about a local build isn't in my channel log at
least.


Regards,

Adam


--
To UNSUBSCRIBE, email to debian-gcc-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 7c2b6196f6684cd2459583398ae93e85@mail.adsl.funky-badger.org">http://lists.debian.org/7c2b6196f6684cd2459583398ae93e85@mail.adsl.funky-badger.org
 
Old 05-08-2012, 05:24 PM
Cyril Brulebois
 
Default Accepted gcc-defaults 1.118 (source all amd64)

Matthias Klose <doko@debian.org> (08/05/2012):
> stop blaming the compiler for unrelated build failures. You did so already
> yesterday:
>
> <mbiebl> awesome, libchamplain ftbfs with an ICE in gcc-4.7
> <mbiebl> yeah for switching the default compiler
> <KiBi> doko: It looks like I'm going to take up your offer. Please fix that.
> <mbiebl> doko: you have your first problem of gcc-4.7 breaking an ongoing
> transition
> <adsb> mbiebl: that might be the gmp10 bug
> <doko> KiBi, can't reproduce the libchamplain ICE locally. could you do a
> local build as well?
>
> and I didn't get a reply on that.

And to be complete, after checking with others, that you didn't get any
answer shouldn't come as a surprise, since nobody received your question.

kibi@Thorfinn:~$ grep -A 50 'that might be the gmp10 bug' irclogs/Debian/#debian-release.log|grep doko
2012-05-07 21:56:13[ doko] KiBi, looking
2012-05-07 22:12:37::: doko [~doko@216.64.156.98] has quit [Read error: Operation timed out]

Mraw,
KiBi.
 

Thread Tools




All times are GMT. The time now is 12:30 PM.

VBulletin, Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.
Content Relevant URLs by vBSEO ©2007, Crawlability, Inc.
Copyright 2007 - 2008, www.linux-archive.org